**1. Introduction**

The useful life of cutting tools substantially increases after the deposition of thin wear-resistant films [1,2]. This is due to the film hardness of about 25–30 GPa [3] exceeding by many times the hardness of the bulk material. Magnetrons [4] and vacuum arc [5] are mainly used to produce the metal vapor needed for coating synthesis. The density of magnetron discharge plasma near the product surface and the sputtering rate are substantially enhanced when pulsed DC magnetrons [6–13] are used. Metal vapor can be also produced due to sputtering a target at the bottom of a hollow cathode [14].

To prevent the formation of metal droplets in the coatings, the vacuum arc plasma can be filtered using magnetic field. An overview of filtered vacuum arc deposition systems based on magnetic ducts is presented in [15]. The droplets can be also removed from the plasma by transformation of radial plasma streams emitted by the arc cathode spots on the side surface of a cylindrical cathode into an axial stream by means of "single bottle neck" magnetic field [16]. Filtered vacuum arc with ion-species-selective bias has been applied to the synthesis of metal-doped diamond-like carbon films [17]. A magnetic island filter comprising three external coils, which generate a uniform magnetic field, and an internal coaxial coil with a cylindrical permanent magnet in its core, placed within the magnetic island, that generate a field in the opposite direction to the external field is described in [18]. Instead of coils, an internal curvilinear spiral was used in [19], which was transparent and allowed observation of the plasma movement from the arc cathode to the substrate. Analysis of the workpiece surface in [20] revealed that its roughness and the number of cathode spots show no direct relation because the current density per cathode spot does not change according to the number of cathode spots. In a pulsed vacuum arc discharge, the microparticles can be charged, and a roughly quadratic dependence of particle charge on the particle diameter was observed [21] with a 1-μm-diameter particle having a positive charge of ≈1000 electronic charges and a 5-μm-diameter particle having

**Citation:** Metel, A.; Melnik, Y.; Mustafaev, E.; Minin, I.; Pivkin, P. Combined Processing of Micro Cutters Using a Beam of Fast Argon Atoms in Plasma. *Coatings* **2021**, *11*, 465. https://doi.org/10.3390/ coatings11040465

Academic Editor: Cecilia Bartuli

Received: 16 March 2021 Accepted: 13 April 2021 Published: 16 April 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

a charge of ≈25,000 electronic charges. Highly adherent CrN films were magnetron sputter deposited in [22] after the filtered cathodic arc etching pretreatment.

Generally, the tools are preliminarily heated and activated. Such pretreatment allows an appreciable improvement of the coating adhesion. More improvements of the adhesion and other coating properties are available when instead of DC biasing, high-voltage pulses are used [23–26].

The thicker the coating, the longer is its useful life. Nevertheless, on roughing tools with the cutting edge radii amounting to ≈15–20 μm, the coating thickness usually does not exceed 5–7 μm. At higher values of the coating thickness, the cutting edge radii grow to inadmissible values exceeding 30 μm, which are characteristic of blunted tools [27]. The 30-μm radius of the cutting edge exceeds the uncut chip thickness of micro-milling. Therefore, the material deformation mechanisms are often dominated by the plowing effect [28], and the quality of the surface is dissatisfactory [29].

The problem of cutting tools blunting caused by the coating deposition is especially acute for micro cutters with a cutting edge radius of less than 10 μm. For instance, cylindrical end-milling micro cutters work in cutting conditions, at which the depth of cut is commensurate with the cutting edge radius. It is impossible to increase the cutting stability during micro-milling through an increase in the depth of cut because of a significant degree of the tool bending. Therefore, keeping the minimum cutting edge radius is very important [30].

The minimum radius of the cutting edge will improve the cutting conditions and increase the likelihood of the cutting into the material of the whole cutting wedge, not just its tip, which can reduce cutting forces [31] and vibration [32], improve the chip formation, use cutting patterns with a minimum depth of cut at a relatively high feed per tooth, and reduce the likelihood of the workpiece abrasion.

To date, a coating with a thickness exceeding 2–3 μm cannot be deposited on micro cutters because after grinding, the radius of the cutting edge is in the range of chip thickness: about 10 μm. The wear-resistant coatings can highly improve the cutting tools' efficiency [33–36]. However, in the case of micro tools, their cutting edge radii will increase by 2–2.5 times, which will not allow a stable cutting process. In this case, an increase in the thickness of the cut chips is impossible, because micro cutters are small and have a lower ability to resist bending, which promises their destruction when bending.

Thus, there is a limitation on the maximum feed per tooth value, and there is a minimum feed per tooth limit by the minimum chip thickness. The latter relates to the minimum allowable cutting edge radius that must be met for a stable cutting process. Chips simply will not be formed if their thickness is less than the minimum. Therefore, the only way to increase the useful life of micro cutters without losing the cutting process stability in conditions of parametric tool failure is to reduce the cutting edge radius.

Experimental investigations regarding the formation of the cutting edge geometry during an immersed tumbling process are presented in [27]. It was found that the processing time is mainly responsible for the size of the cutting edge radius. The lapping medium defines the chipping of the edge. During the experiments, micro cutters were manufactured with a cutting edge radius of ≈4.0 μm, which is lower than the values available with traditional grinding methods. Another method for diminishing the cutting edge radius based on the ELID (electrolytic in-process dressing) technique allows manufacturing cutters with the radius of approximately 3 μm [37]. However, it cannot be applied for small-sized tools.

We propose in the present study to reduce the cutting edge radius by etching the tool surface with a beam of fast argon atoms [38–41]. The etching will allow reduction of the coated tool radius lower than the radius of the initial tool.

#### **2. Experimental**

We processed end mills by Cerin (IT) under catalog number 105.030031240 with the nominal diameter of the working part amounting to 3 mm (h10) and length of 12 mm as well as the nominal angle of inclination of the chip groove equal to 30 degrees for machining light alloys, aluminum, and titanium (Figure 1). The material of the end mills is solid carbide type HM CK10-20-MG Micrograin: monocarbide (WC-94%; Co-6%) with the carbide grain size of 0.5–0.8 μm, hardness Rockwell 91.8 HRA, and transverse rupture strength 3000 N/mm2. The work surfaces of the end mills are polished and uncoated. According to the catalog, their application areas are light alloys, plastics, reinforced plastics, and titanium alloys.

**Figure 1.** Photograph of a 3-mm-diameter end mill with a 30-degree angle of the chip groove inclination.

While studying the bombardment by a beam of fast argon atoms, it is necessary to compare the initial geometric parameters of the cutting tools with the parameters obtained after processing. Analysis of the changes in the controlled parameters will make it possible to conclude the effectiveness of the fast atom beam. The controlled parameters were the profiles of the front surface located at different distances from the end of the cutter, as well as the angular parameters of the cutting edge, the front, and rear corners, and the cutting edge radius.

During the study, three samples of this type of end mill were measured. To control the initial geometric parameters, the algorithms of the Walter Helicheck Plus (Walter, Germany) measuring system was used, which are based on the use of machine vision methods. The measuring was carried out with cameras of transmitted and reflected light. Measurement inaccuracy by positioning stability for measuring diametrical and linear parameters does not exceed 0.32 μm. The measurement was carried out in a non-contact manner using a CCD (Charge-Coupled Device) matrix camera of transmitted light and a CCD camera of reflected light with ×400 magnification.

A hydraulic chuck with an SK 50 tool cone was used for fixing the measured samples of cutters. The obtained results of primary measurements for the working geometry of three samples are presented in Table 1.


**Table 1.** Results of primary measurements for the working geometry of three samples.

The main geometrical parameters of a solid carbide end micro mill and radius of cutting on the radial section are shown in Figure 2. The main geometric parameters of the cutting edge were measured in three areas: A, B, and C. In the range of each area, the cutting edge radius was controlled in sets of sections: *reA*1, *reA*2, *reA*3, and *reAn* for A area, *reB*1, *reB*2, *reB*3, and *reBn* for B, and *reC*1, *reC*2, *reC*3, and *reCn* for C.

**Figure 2.** Main geometrical parameters of a solid carbide end micro mill.

Cutting edge radius was measured with an optical 3D measuring system MicroCAD Premium plus. Its measuring algorithm is to measure a strip of light using a micromirror projector. Figure 3a shows a photograph of the measuring system and Figure 3b elucidates the measurement of the cutting edge radius.

**Figure 3.** Optical measuring system MikroCAD premium+ manufactured by GFMesstechnik GmbH (**a**) and a 3-mm-diameter end mill installed under the system (**b**).

During the measurements, the radius of the cutting edge was estimated in four radial sections distanced from each other at 0.05 mm and distanced from the mill end at 2 mm (Figure 4), 7 mm, or 12 mm.

**Figure 4.** The end mill tooth at a distance of 7 mm from the mill end. (**a**) Image from optical 3D scanning camera; (**b**) topography of the cutting edge.

The radius of the cutting edge was measured in three areas: A, B, and C, which were located from each other with an equal step of 5 mm at a distance of 2, 7, and 12 mm. In each of the areas, the value of the radius *reA*, *reB*, and *reC* is found as the arithmetic mean of the measured values for four sections located at 0.05 mm distance from each other (Equations (1)–(3)).

$$r\_{\varepsilon A} = \frac{r\_{\varepsilon A1} + r\_{\varepsilon A2} + r\_{\varepsilon A3} + \dots + r\_{\varepsilon An}}{n} \tag{1}$$

$$r\_{\varepsilon B} = \frac{r\_{\varepsilon B1} + r\_{\varepsilon B2} + r\_{\varepsilon B3} + \dots + r\_{\varepsilon Bn}}{n} \tag{2}$$

$$r\_{\rm rC} = \frac{r\_{\rm cC1} + r\_{\rm cC2} + r\_{\rm cC3} + \dots + r\_{\rm cCn}}{n} \tag{3}$$

The radius of the cutting edge was measured for the basic type of end mill with grinding clearance and rake surfaces. The average value of the cutting edge radius per 2-mm-distanced area (A) was equal to *reA* 10.8 μm (Figure 5); the average value of the cutting edge radius per 7-mm-distanced area (B) was equal to *reB* 10.5 μm; and the average value of the cutting edge radius per 12-mm-distanced area (C) was equal to *reC* 10.25 μm.

**Figure 5.** Cutting edge radii *reA*1, *reA*2, *reA*3, and *reA*<sup>4</sup> in four sections distanced from the mill end at 2 mm and distanced from each other at 0.05 mm.

The complete experimental process of the present work comprises:


### **3. Etching the End Mills with Fast Atoms**

The experimental system for processing the end mills is presented in Figure 6. A 20-cm-diameter concave grid with a surface curvature radius of 20 cm is fixed to the high-voltage feedthrough in the center of the vacuum chamber.

**Figure 6.** Scheme of a beam generation by a grid immersed in plasma.

On the grid surface, holes with a diameter of 7 mm are evenly distributed at a distance of 8 mm between their centers. At a distance of 6 cm from the chamber wall is installed a rotating holder for the tools being processed by the beam. The axis of the holder rotation passes through the focal point of the concave grid surface.

At an argon pressure in the chamber *p* ≈ 0.5 Pa, an increase in the voltage between the anode and the chamber to several hundred volts leads to establishing a gas discharge with a current in the anode circuit *I*<sup>a</sup> up to 4 A and a discharge voltage of *U*<sup>d</sup> = 400–500 V [42]. The chamber is filled with a brightly glowing homogeneous plasma, which is separated from its walls by a cathode sheath and from the grid by a grid sheath of the ion space charge. The sheath width *d* can be calculated using the derived from the Child–Langmuir law [43] following expression

$$d = (2/3)\varepsilon\_{\rm o}^{-1/2} (2\varepsilon/M)^{1/4} (\mathcal{U} + \mathcal{U}\_{\rm d})^{3/4} / j^{1/2},\tag{4}$$

where <sup>ε</sup><sup>o</sup> is the electric constant equal to 0.885 × <sup>10</sup>−<sup>11</sup> F/m, *<sup>e</sup>* is the electron charge, *<sup>M</sup>* is the ion mass, *U* is the grid bias voltage, and *j* is the ion current density. When the ion mass *<sup>M</sup>* <sup>=</sup> *AMA* is measured in atomic mass units *<sup>A</sup>* = 1.66 × <sup>10</sup>−<sup>27</sup> kg, the sheath width is equal to

$$d = 2.34 \times 10^{-4} (M\_A)^{-1/4} (\mathcal{U} + \mathcal{U}\_c)^{3/4} / \dot{\jmath}^{1/2}.\tag{5}$$

When *U* = 0, the chamber and the grid are equipotential, and both sheaths are of the same width. At *p* ranging from 0.2 to 1 Pa, the voltage *U*<sup>d</sup> between the chamber and the anode is virtually independent of the argon pressure *p* at a constant current *I*a in the anode circuit. However, at *p* < 0.2 Pa, a decrease in pressure causes an increase in the discharge voltage *U*d, and at *p* = 0.02 Pa, it reaches a value of *U*<sup>d</sup> ≈ 1 kV. With an increase in the voltage *U* between the chamber and the grid from zero to 5 kV, the current *I* in the grid circuit at a constant current *I*a in the anode circuit approximately doubles, the width of the sheath *d* between the plasma and the grid grows to 5–10 cm, and the discharge voltage *U*<sup>d</sup> decreases approximately by two times.

The density of the gas atoms at room temperature and pressure *p* = 0.02 Pa amounts to *<sup>n</sup>* = 0.5 × 1019 <sup>m</sup>−<sup>3</sup> [44], and the average path of argon ions between charge exchange collisions, called the charge exchange length, is equal to λ = 1/*n*σ = 1 m. We took into account that for argon ions with an energy of 5 keV, the charge exchange cross-section is equal to <sup>σ</sup> = 2 × <sup>10</sup>−<sup>19</sup> m2 [45,46].

At a pressure of *p* = 0.02 Pa, the width of the grid sheath *d* ≈ 0.05 m is much lower than the charge exchange length λ, and therefore, no formation of fast atoms in the sheath occurs. All ions extracted from the plasma bombard the grid and cause the emission from its surface of the secondary electrons [47]. Hence, the grid emits only two beams of electrons with energy *e*(*U* + *U*d), propagating in opposite directions.

With a pressure increase to 0.2 Pa, accelerated ions turn in the sheath into fast atoms escaping from the sheath. The energy of each fast atom, *e*ϕ, corresponds to the potential ϕ of the point in the sheath where it appears. With a pressure increase from 0.2 to 2 Pa, the number of fast atoms increases tenfold, their energy decreases from 5 to 500 eV, and the neutral beam current grows up.

For an estimate of the beam diameter, the distribution of the etching rate by fast atoms of a polished titanium target covered with a mask on its surface was measured. The straight boundary between them is located horizontally (Figure 7). The target was etched with fast atoms, and then, the height of the step between the covered by the mask and open substrate surfaces was measured along the border between them using a mechanical profiler Dektak XT. The etching rate distribution at various distances *Z* between the target and the grid revealed the beam diameter dependence on *Z*. As *Z* increases, the diameter *D* decreases from 45 mm at *Z* = 17 cm to 6 mm at *Z* = 20 cm, and as the distance increases from *Z* = 21 cm to *Z* = 24 cm, the diameter *D* increases from 7 mm up to 52 mm.

**Figure 7.** Schematic of the step formation on a target at the distance *Z* from the grid due to etching with fast argon atoms.

Figure 8a presents a photograph of the accelerating grid fastened to the high-voltage feedthrough and the end mill placed on the rotating holder. The grid was distanced at 22 cm from the axis of the rotating holder where the beam diameter amounted to 2 cm. For this reason, the 12-mm-long cutting part of the end mill was etched quite homogeneously by the fast atoms. The beam of fast atoms was generated by the grid immersed in the argon plasma and negatively biased to 5000 V (Figure 8b).

**Figure 8.** Photographs of the accelerating grid and 3-mm-diameter end mill on the rotating holder (**a**) as well as the accelerating grid immersed in plasma and generating a fast atom beam (**b**).

The results of the mills etching are presented in Table 2. They show that at the beginning, for the sections distanced from the mill end at 7 mm, the etching rate is maximum and amounts to 3 μm/h. Further on, the etching rate falls down, and after a 3-h-long etching, it is close to zero for the section distant from the mill end at 7 mm and to 0.2 μm/h for the sections distant from the mill end at 2 and 12 mm.



The average value of the cutting edge radius after etching per 2-mm-distanced area (A) was equal to *reA* 4.25 μm; the average value of the cutting edge radius per 7-mm-distanced area (B) was equal to *reB* 3.65 μm (Figure 9); the average value of the cutting edge radius per 12-mm-distanced area (C) was equal to *reC* 4.1 μm.

**Figure 9.** Profiles of the cutting edge radii after etching in four sections distanced from the mill end at 7 mm and distanced from each other at 0.05 mm.

The analysis of the entire set of measured cutting edge radii showed that the best results were obtained for the 3-h-long etching.

A further increase in the etching time had no benefit for the tool sharpening. It should be mentioned that after quite long etching of the end mills, no significant increase in the surface roughness has been observed (Figure 10).

**Figure 10.** Measurements of the end mill roughness before (**a**) and after the etching (**b**).

After a 3-h-long etching of the end mill, a wear-resistant coating was synthesized on its surface. The synthesis was carried out according to the standard technology on a Platit π 311 system for the deposition of wear-resistant coatings manufactured by Platit (Switzerland). The deposited diamond-like coating (DLC) was a two-layer composition: an adhesive sublayer based on a complex nitride (CrAlSi)N and an outer wear-resistant DLC layer (Figure 10b). The choice of this particular coating (CrAlSi)N/DLC as an object for research is not accidental. This compound can be used to improve the performance of hard-alloy tools, including small-sized ones. To estimate the thickness of the synthesized coating, a cylindrical mask with an inner diameter of 3 mm was preliminary put on the end mill.

Within an hour, a DLC coating was synthesized on the end mill and its mask. After removing them from the chamber, the mask was detached from the end mill. Using a Dektak XT mechanical profilometer, the height of the step between the coating surface and the end mill surface screened with the mask was measured (Figure 11a). The thickness of the wear-resistant coating synthesized on the end mill is equal to the step height of 2.4 μm.

**Figure 11.** Profilogram of the end mill surface without the detached mask (**a**) and construction of a two-layer DLC coating deposited on a 3-mm-diameter end mill (**b**).

Figure 12 shows the experimentally obtained measurement results, which were performed in four sections of the cutting edge of a 3-mm diameter end mill after a 3-h-long etching and deposition of a two-layer DLC coating. The average value of the cutting edge radii after etching and deposition of a two-layer DLC coating per 2-mm-distanced area (A) was equal to *reA* 6.25 μm; the average value of the cutting edge radius per 7-mm-distanced area (B) was equal to *reB* 5.5 μm; and the average value of the cutting edge radius per 12-mm-distanced area (C) was equal to *reC* 6.2 μm (Figure 12) (Table 3).

**Figure 12.** Profiles of cutting edge radii after etching and deposition of a two-layer DLC coating in four sections distanced from the mill end at 12 mm and distanced from each other at 0.05 mm.



Figure 13 presents SEM images (magnification 5000 times) of the end mill cutting edges after the diamond-like coatings deposition on the end mills sharpened through grinding (a) and by fast argon atoms (b).

**Figure 13.** SEM images of the cutting edges coated with DLC films after sharpening through grinding (**a**) and by fast argon atoms (**b**).
